Prediksi Penjualan Hijab menggunakan Metode Extreme Learning Machine (ELM) (Studi Kasus: Vie Hijab Store)

Kevin Nadio Dwi Putra (Fakultas Ilmu Komputer, Universitas Brawijaya)
Muhammad Tanzil Furqon (Fakultas Ilmu Komputer, Universitas Brawijaya)
Novanto Yudistira (Fakultas Ilmu Komputer, Universitas Brawijaya)

Jurnal Pengembangan Teknologi Informasi dan Ilmu Komputer
Vol 4 No 6 (2020): Juni 2020

In the industrial world nowadays, many people have established any business from micro to macro. One of the business that put focus of this research is a home industry selling hijab named Vie Hijab Store. Vie Hijab Store has increased sales year by year, but there are problems in supply management of raw materials in the form of fabrics. Therefore, by predicting the amount of hijab sales, it is expected to be able to assist the owner with building consideration in making the decision to purchase raw material in a certain period. This research will use Extreme Learning Machine (ELM) prediction method which has advantages in learning speed and for calculating the error rate of the predicted results using Mean Average Percentage Error (MAPE). The smallest MAPE results obtained for the Khimar model were 22% and for the Pashmina model by 12% using 4 features, 5 hidden nodes, a binary sigmoid activation function, and a data ratio of 60%: 40% for the Khimar model and 70%: 30% for the Pashmina model.